Expected liberalizations in the Medicare requirements for risk contracting will permit Provider Service Organizations (PSOs) to contract with Medicare without prior experience, while relying on complex risk sharing arrangements with providers in lieu of meeting capital and surplus requirements. Hundreds of new PSO plans are expected to be organized to take advantage of the new opportunities. Many of these new plans will involve a large number of interlocking subcapitation and risk sharing arrangements with hospitals, primary care physicians and specialists that can not be modeled with existing commercial products. To fill this need, and provide a sound basis for designing PSO based risk contracts for Medicare, a comprehensive actuarial rate model will be developed from actual Medicare data, with the capacities needed to determine the cost impact of specific subcapitation and risk sharing arrangements. Unique features of the new rate structure include: (1) derived from actual Medicare FFS and HMO claims data, (2) provide capacity to model capitation and other risk sharing arrangements, (4) model the effect of carving out specific professional procedures or hospital discharge diagnoses, (4) capacity to model managed care savings and substitution effects explicitly by type of provider, and (5) incorporate multiple stop loss reinsurance arrangements that affect limited sets of providers and services. PROPOSED COMMERCIAL APPLICATION: Hundreds of Provider Service Organization (PSO) health plans are expected to be formed in response to liberalizations in qualifications of health plans that offer risk contracts in Medicare. The reduced requirements-esp, the reliance on complex capitation and other risk sharing arrangements among providers, the limited capitalization requirements, the absence of prior experience and sole reliance on Medicare contracts-will require very different actuarial tools than are currently available. In addition to PSOs, HMOs and other insurers provide potential markets for the kind of detail and analytical capacities planned for the models.